User features

User custom data

Overview

This feature allow third party apps to save custom data into user database

Properties

Field	Type	Description
data_key	String(92)	Data key, can be a path with dot separation
data_group	String(92)	Feature group name
data_value	jsonb	Json data
user_id	int64	price per item in quantity.

APIs

VERD	Endpoint	Parameters	Data	Response	Desc
GET	/api/user-data/get-one-by-key	userId, dataKey		UserData	Find a data key for specific user, throw HTTP 404 if not found
POST	/api/user-data/find-one-by-key		UserDataKeyDto	UserData	Find a data key for specific user
POST	/api/user-data/find-many-by-group		UserDataGroupDto	UserData[]	Find many data by group for specific user
POST	/api/user-data/save-one-by-key/:userId/:dataKey		UserDataSaveDto	UserData	Save or update with data key for specific user

export class UserDataKeyDto {
  // Data key, no need to unique, (userId, dataKey) must be unique
  dataKey: string;

  // User id
  userId: string;
}

export class UserDataGroupDto {
  // Data key, no need to unique
  dataGroup: string;

  // User id
  userId: string;
}

export class UserDataValueDto {
  // Data key, no need to unique, (userId, dataKey) must be unique
  dataKey: string;

  // User id
  userId: string;

  // Full update, previous data will be overwrited
  dataValue?: any;

  // Partial update, only update provide fields, other fields is untouched
  partialDataValue?: any;
}

Usage examples

1. Store data for a game/app for user id 'user1' when user is first created

Call POST /api/user-data/save-one-by-key/

With data

{
  "userId": "user1",
  "dataGroup": "basic",
  "dataKey": "user-profile",
  "dataValue": {
    "level": 1,
    "displayName": "Hello"
  }
}

2. Later update some attribute as user progress

Call POST /api/user-data/save-one-by-key/

With data

{
  "userId": "user1",
  "dataKey": "user-profile",
  "partialDataValue": {
    "level": 2
  }
}

3. Retreive data to display on UI

Call GET /api/user-data/get-one-by-key/user1/user-profile

Response data

{
  "userId": "user1",
  "dataKey": "user-profile",
  "dataGroup": "basic",
  "dataValue": {
    "level": 2,
    "displayName": "Hello"
  }
  //... other data
}

4. Retreive all data from a feature group

Call GET /api/user-data/find-many-by-group

With data

{
  "userId": "user1",
  "dataGroup": "basic"
}

Response data

[
  {
    "userId": "user1",
    "dataGroup": "basic",
    "dataKey": "user-profile",
    "dataValue": {
      "level": 2,
      "displayName": "Hello"
    }
    //... other data
  },
  {
    "userId": "user1",
    "dataGroup": "basic"
    //... other data
  }
]

Tips and tricks

• Try to group data as must as possiple to reduce records generation which will impact performance
• Try to make data value small enough for a single atomic feature, since update method is not thread safe, if store very large object into as single data will cause data inconsistancy for highload environment
• Summary, how to break and how to merge data need to be balanced